I guess I don't really do much 'real' bioinformatics anymore, but its been a while since I had to parse any biological data. It seems that the proliferation of web services and the slow uptake of some standards for data sharing has actually had some positive effects.
The last few bioinformatics-related projects I did included the use of a) NCBI E-utils web services (no parsing required because it goes directly into a java object) and the protein data from beta.uniprot.org (not parsing required because the data is retrieved in RDF which tools like Jena can parse automatically).
Personally I think the days of parsing text files are on the way out. Coincidentally I think the term 'bioinformatics' will fall out of favor and be replaced by 'biology' on the one hand and 'information science' on the other.